[11330] in www-talk@info.cern.ch
Re: SV: distinguising user- from browser-generated requests
daemon@ATHENA.MIT.EDU (Al Gilman)
Wed Jun 23 13:14:43 1999
Date: Wed, 23 Jun 1999 13:19:58 -0400
To: "Christian Bxhn" <chrb@online.no>, "Nicholas Kushmerick" <nick@ucd.ie>
From: Al Gilman <asgilman@iamdigex.net>
Cc: "WWW-TALK" <www-talk@w3.org>
In-Reply-To: <01bebd97$5108dfc0$LocalHost@bohns>
Resent-From: www-talk@w3.org
Well, the referer URI will not discriminate between the two cases that Ni=
ck
asked about.
Nick should look into the work of the Web Characterization Activity
<www.w3.org/WCA/> for other people working on questions such as this: wha=
t
one can learn from what one can log.
Al
At 06:42 PM 6/23/99 +0200, Christian B=F8hn wrote:
>Check out the referer header, it should referer the page which contained
it, but of course this is not a safe method.
>
>HTH,
>Christian B=F8hn
>
>
>
>>When an HTML page is rendered, the browser might spawn additional reque=
sts=20
>>to the server (images, frames, style sheets, applets, ...). Within HTT=
P,=20
>>can the server reliably distinguish such "browser-generated" requests f=
rom=20
>>actual user requests (ie, when a user clicks on a link)?
>>
>>To make this concrete, suppose an HTML page contains:
>> <A href=3D"q.jpg"><IMG src=3D"q.jpg"></A>
>>Can the server distinguish between the browser's request for "q.jpg" to=
=20
>>render the HTML, and
>>the user's request for "q.jgp" by clicking on the hyperlink?
>>
>>Assume that the HTML is not under my control -- ie, I can't simply repl=
ace=20
>>the href above with "q.jpg?user".
>>
>>thanks for any suggestions!
>>
>>-- Nick
>>
>>- + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - +
>>+ Nicholas Kushmerick -
>>- Department of Computer Science, University College Dublin +
>>+ nick@ucd.ie www.cs.ucd.ie/staff/nick -
>>- + - + - + - + - + - + - + - + - + - + - + - + - + - + - + - +
>>
>>
>=20
